Research map / 3D vision / NeRF and radiance fields
Dynamic Neural Radiance Fields for Monocular 4D Facial Avatar Reconstruction
Guy Gafni, Justus Thies, Michael Zollhofer, Matthias Niessner
Check your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use the full dynamic neural radiance field instead of DVP | Facial reenactment · Self-reenactment test set | LPIPS: 0.10 -> 0.06 | Improved |
| Add background image input (ablation) | Dynamic NeRF · Self-reenactment test set | LPIPS: 0.18 -> 0.06 | Improved |
| Add conditioning on facial dynamics (ablation) | Dynamic NeRF · Self-reenactment test set | LPIPS: 0.11 -> 0.06 | Improved |
| Train on the full sequence rather than 50% of it (ablation) | Dynamic NeRF · Self-reenactment test set | LPIPS: 0.07 -> 0.06 | Improved |
| Train on the full sequence rather than 25% of it (ablation) | Dynamic NeRF · Self-reenactment test set | LPIPS: 0.09 -> 0.06 | Improved |
| Add learnable per-frame latent codes (ablation) | Dynamic NeRF · Held-out generated images | LPIPS: 0.068 -> 0.059 | Improved |
Closest papers by meaning
- Learning Compositional Radiance Fields of Dynamic Human Heads (CVPR 2021 · similarity 0.92)
- Learning Dynamic Facial Radiance Fields for Few-Shot Talking Head Synthesis (ECCV 2022 · similarity 0.90)
- High-Fidelity Facial Avatar Reconstruction From Monocular Video With Generative Priors (CVPR 2023 · similarity 0.90)
- Pixel-Aligned Volumetric Avatars (CVPR 2021 · similarity 0.90)
- Artist-Friendly Relightable and Animatable Neural Heads (CVPR 2024 · similarity 0.89)
- High-Fidelity 3D Head Avatars Reconstruction through Spatially-Varying Expression Conditioned Neural Radiance Field (AAAI 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
